jurgenaut23

9 ms·

91.
Learn in Infinite Dimensions(neuraloperator.github.io)
2 points by jurgenaut23 3y ago|2 comments
92.
by jurgenaut23 3y ago
Do you guys have experience with neural operators and this library in particular? Any cool projects that you could showcase here? I got interested in this because of this recent paper in Science: https://www.science.org/doi&
93.
by jurgenaut23 3y ago
Do you have any tips for self-publishing, or reliable resources that talk about it? Or do you manage PR, did you get access to any form of media coverage? I thought this was the main point of getting a publisher.
94.
by jurgenaut23 3y ago
One good test to see if you're doing it right, by the way: do these things with people who devote most of their time and energy to it, and see if you can still enjoy yourself. If so, you're doing it right ;-)
95.
by jurgenaut23 3y ago
I don't believe that you can only be good at a couple things. You can be _great_ at only a couple things. And you can only be the _best_ at a single thing. As it happens, I think you can actually be fairly good at _many_ things, which
96.
by jurgenaut23 3y ago
I can't help to wonder how the bitrate would vary with how cooked they are. Finally one way to settle the debate about what's really "al dente".
97.
by jurgenaut23 3y ago
If you read "Limits to Growth" as if it were some contemporary influencer or futurist's content, this might indeed be disappointing. But the truth is, "Limits to Growth" is one of the first attempt to articulate the
98.
by jurgenaut23 3y ago
You’re a statistician, aren’t you?
99.
by jurgenaut23 3y ago
Ah, didn't know about it, but that's exactly my thought.
100.
by jurgenaut23 3y ago
Well, I think that, despite being a joke, your comment is deeper than it looks like. As model capabilities increase, the likelihood that they interfere with the instructions that we provide increases as well. It’s really like hiring someone
101.
by jurgenaut23 3y ago
This reminds me that, with our species, the real power always lies with those that tell the better story. Skynet and HAL 9000 had a more profound impact on those researchers than years of practice and study in the field. No surprise therefo
102.
by jurgenaut23 3y ago
Agreed, this was my first impression as well. I spend my whole day in front of much more complex code, but trying to put myself in the shoes of a young kid, I would be really misled by this (unnecessary) noise.
103.
by jurgenaut23 3y ago
I predict this isn't the last episode of this amazing soap opera.
104.
by jurgenaut23 3y ago
I think that the non-profit status of OpenAI was ultimately its demise as well: as the stakes get higher, people just cannot help themselves but to get (too) interested in more than just the original mission. Being a non-profit doesn't
105.
by jurgenaut23 3y ago
If I weren't so adverse to conspiracy theories, I would think that this is all a big "coup" by Microsoft: Ilya conspired with Microsoft and Altman to get him fired by the board, just to make it easy for Microsoft to hire him
106.
by jurgenaut23 3y ago
OK, I was part of those that thought that Altman did something wrong and the board was trying to do some damage control, but it actually seems that the board is simply out of their mind. This was _indeed_ another possibility, but seemed muc
107.
by jurgenaut23 3y ago
No, this passage tells me that the board wants to cover their ass: "he was not consistently candid in his communications with the board [...]. The board no longer has confidence in his ability to continue leading OpenAI." It'
108.
by jurgenaut23 3y ago
Well, maybe not?
109.
by jurgenaut23 3y ago
Yes, very likely Altman has done something _very_ wrong, and the board wants to maintain plausible deniability.
110.
by jurgenaut23 3y ago
Ouch, this must have been a _very_ disagreeable departure, or the PR department had some really deep resentment towards Altman. I haven't seen such an abrupt and brutal announcement for quite a while.
111.
by jurgenaut23 3y ago
Well, it shows also how crazy difficult that is.
112.
by jurgenaut23 3y ago
I don’t even need to read the article to know that this is BS. They ran on renewables for their electricity, which is likely less than 50% of their _energy_ consumption. As long as journalists don’t know the difference, we won’t be in a goo
113.
by jurgenaut23 3y ago
Great analogy of the wider world, by the way.
114.
by jurgenaut23 3y ago
I don't drink alcohol at all and I have noticed this trend as well, at least where I live (Switzerland).
115.
by jurgenaut23 3y ago
Ouch, that hurts. But indeed, that shows that most people have the attention span of a gold fish.
116.
by jurgenaut23 3y ago
TLDR; the guy didn't pass Toptal Codility test.
117.
by jurgenaut23 3y ago
Thank you so much for investigating this matter so thoroughly and fighting for justice so eagerly. As the father of a 4-month-old baby, I couldn't imagine anything more devastating than having him taken away by the authorities for some
118.
by jurgenaut23 3y ago
Unless we live in orthogonal dimensions, you know that free unlimited energy is never devoted to recycling. Plus, fusion will produce electricity, which needs to be stored in batteries. Or it will be used to produce hydrogen, which in turn
119.
by jurgenaut23 3y ago
Please. I know lots of folks here live in Silicon Valley and are breastfed with this kind of fantasy, but making our species multi-planetary isn't just a matter of energy availability. Plus, and most importantly, trashing our one and o
120.
by jurgenaut23 3y ago
The issue with free cheap energy such as fusion isn't its direct byproducts, but its impact on society. It will fuel an economic growth without precedent, which in turn will increase the consumption of material goods, which in turn wil

More ›